ESTIMATING WHEN EVENTS OCCURRED

拟议的研究涉及的一个重要方面， 自传体记忆--记住事件发生时间的能力。 很明显，一个保存了时间的纪念记录 一个人的经验的连续性是智力的核心。 以及对自我的连贯看法。然而， 关于时间记忆的理论问题还没有被探索， 系统的实证研究尚待开展。可用数据 表明，虽然事件发生的时间被保存在 记忆在某种程度上，有大量的不准确性和偏见， 临时报告。在拟议的研究中，我们的目标是探索和解释 观察到时间报告中的偏差。 下面概述的研究来自我们最近的工作， 在有序域中的存储器中的分层组织的结果， 包括时间记忆时间被划分成各种大小的单位 （小时、天、周）。当信息在细节层面上不准确时 为了回答问题，来自较大时间单位的信息可能 影响估计，导致报告出现偏差。初步证据 表明偏见是记忆不准确的结果， 随着时间的推移而增加。我们建立了两个定量模型， 指定不同详细程度的信息被 结合在一起产生时间报告。我们建议推行一系列 研究提供了系统的信息，不准确和偏见， 报告.当前的目标是检查模型是否充分考虑到 观察到的偏差。解释偏见的来源将有助于我们 理解时态信息如何被表示的理论目标 在记忆中，以及涉及的检索和估计过程， 回答时间问题它也可能有助于实际的 发展方法收集更准确的社会和健康数据的问题 涉及临时报告的相关数据。

The proposed research is concerned with an important aspect of autobiographical memory - the ability to remember when events occurred. It seems clear that a memorial record which preserves the temporal continuity of a person's experience is central to intellectual functioning as well as to a coherent view of the self. Yet major theoretical issues concerning temporal memory have not been explored, and systematic empirical studies remain to be carried out. Available data indicate that, while the time of occurrence of events is preserved in memory to some extent, there is substantial inexactness and bias in temporal reports. In the proposed research, we aim to explore and explain observed biases in temporal reports. The studies outlined below emerge from our recent work on the consequences of hierarchical organization in memory in ordered domains, including temporal memory. Time is divided into units of various sizes (hours, days, weeks). When information is inexact at the level of detail required to answer a question, information from larger temporal units may affect estimation, leading to bias in reporting. Preliminary evidence suggests that bias is a consequence of inexactness in memory and increases over time. We have developed two quantitative models which specify ways in which information at different levels of detail is combined in producing temporal reports. We propose to carry out a set of studies which provide systematic information on inexactness and bias in reporting. The immediate goal is to examine if the models fully account for observed bias. Explaining the sources of bias will contribute to our theoretical goal of understanding how temporal information is represented in memory, and of the retrieval and estimation processes involved in answering temporal questions. It may also contribute to the practical problem of developing methods to collect more accurate social and health related data involving temporal reports.